ARTEMIS: a self-managed app that helps adults lose weight – without clinician input

Weight management programmes work. But delivering them at the scale needed – to the millions of people who could benefit – remains one of the hardest challenges in public health. The most effective digital programmes still rely on intensive support from a coach or clinician, which drives up costs and limits reach. The question is whether a fully self-managed tool, requiring no professional input at all, can deliver clinically meaningful results.

ARTEMIS set out to answer that question.

Our approach and partners

ARTEMIS – Adults Regulating Their weight Everyday with Mobile Internet Support – is a theory-based mobile app designed to support weight loss through self-regulation. Users weigh themselves daily, set evidence-based action plans, track their weight, and complete weekly reflections. There is no coach, no clinician, no appointments.

The team evaluated ARTEMIS in a large pragmatic randomised controlled trial of 1,607 adults living with obesity, recruited through digital channels. Half were given access to the app; half received advice alone. Participants were followed for six months. The entire trial – recruitment, consent, data collection, follow-up – was conducted remotely, testing whether this approach could work at the scale a real-world rollout would demand. NIHR ARC funding was essential, supporting the PhD studentship that developed the initial intervention, the early app build with a software engineer, and the full trial.

What we found – and why it matters

The trial produced strong results:

1. Greater weight loss. Participants using ARTEMIS lost 1.85kg more than those receiving advice alone.
2. More than double the odds of meaningful weight loss. Users were over twice as likely to achieve 5% or more body weight loss at six months – the threshold considered clinically meaningful.
3. No adverse effects. The app reduced symptoms of disordered eating and improved body satisfaction, demonstrating effectiveness without harm.
4. Equitable outcomes. Results were consistent across age, sex, educational attainment, employment status, deprivation, and ethnicity – meaning the intervention is unlikely to widen existing health inequalities.

Because ARTEMIS requires no clinician time and minimal infrastructure, even modest uptake could translate into meaningful population-level benefit without placing further demand on health services.

What this means

ARTEMIS offers something the NHS currently lacks: a scalable, low-cost weight loss intervention that works, causes no harm, and reaches people equitably regardless of background. It could be offered through healthcare providers, made available through public health programmes, or accessed directly by individuals. For a health system under pressure to shift from treatment to prevention and from analogue to digital, this is the kind of tool that could make that ambition practical.

What needs to happen next

The evidence is strong. The priority now is implementation. The team is working with Oxford University Innovation to identify pathways for making the app available – whether through NHS commissioning, public health programmes, or private routes. Initial conversations with national stakeholders have generated interest in exploring population-level deployment.

Key challenges remain. Digital exclusion means not everyone can access app-based interventions, and regulatory processes for health apps can be slow. Sustaining user engagement over the long term will need further attention as implementation pathways develop.

Lessons for future research

Conducting a trial entirely remotely brought clear advantages – rapid recruitment, a large sample, and evaluation in real-world conditions – but also revealed important challenges:

• Retention was lower than anticipated. Only 40% of participants completed six-month follow-up, typical for digital intervention studies but lower than in trials with face-to-face contact. Novel strategies for maintaining engagement in remote trials are needed.
• Social media recruitment was fast but not diverse. The sample was predominantly white and female, broadly reflecting users of weight loss programmes but limiting generalisability. Complementary recruitment strategies should be planned from the outset for studies targeting under-represented populations.
• Self-reported outcomes require caution. Future evaluations should consider incorporating objective measures where feasible.
• Robust technical infrastructure is essential. Managing platform issues and troubleshooting required substantial effort throughout the trial.